AI/rocm-systems
2
0
Fork 0
Código Issues Pull Requests Ações Pacotes Projetos Versões Wiki Atividade

MI300 counters support for rocprof and rocprofv2(Accumulation from all xccs)

Ver Código-Fonte 
1. Xml files updated for gfx940 counters
2. File plugin changes to allow rocprofv2 backward compatibility for results.csv
3. Changes in rocprofv2 script to use tblextr.py, to generate results.csv just like rocprof

Change-Id: I7798f4411ce01f6fbfffb126de654ed806ca7045
(cherry picked from commit 86cbaf38c436be876f0426fa27803b1e64d90378)


[ROCm/rocprofiler commit: 8f82ff6a46]
Esse commit está contido em:
Saurabh Verma
2023-04-20 20:23:41 -05:00
commit de Ammar ELWazir
pai c61b174908
commit dcd5f1a397
8 arquivos alterados com 503 adições e 46 exclusões
Baixar Arquivo de Patch Baixar Arquivo de Diff Expandir todos os arquivos Colapsar todos os arquivos
projects/rocprofiler/src/core/counters/metrics/basic_counters.xml
+6
Ver Arquivo
@@ -498,6 +498,7 @@
<metric name="SQ_WAVES_RESTORED" block=SQ event=185 descr="Count number of context-restored waves sent to SQs. (per-simd, emulated, global)"></metric>
<metric name="SQ_WAVES_SAVED" block=SQ event=186 descr="Count number of context-saved waves. (per-simd, emulated, global)"></metric>
<metric name="SQ_INSTS_SMEM_NORM" block=SQ event=187 descr="Number of SMEM instructions issued normalized to match smem_level (*2 load/store; *2 atomic; *2 memtime; *4 wb/inv). (per-simd, emulated)"></metric>
<metric name="SQC_ICACHE_INPUT_VALID_READYB" block=SQ event=257 descr=" Input stalled by SQC (per-SQ, nondeterministic, unwindowed)"></metric>
<metric name="SQC_DCACHE_INPUT_VALID_READYB" block=SQ event=260 descr="Input stalled by SQC (per-SQ, nondeterministic, unwindowed)"></metric>
<metric name="SQC_TC_REQ" block=SQ event=262 descr="Total number of TC requests that were issued by instruction and constant caches. (No-Masking, nondeterministic)"></metric>
<metric name="SQC_TC_INST_REQ" block=SQ event=263 descr="Number of insruction requests to the TC (No-Masking, nondeterministic)"></metric>
@@ -505,10 +506,12 @@
<metric name="SQC_TC_DATA_WRITE_REQ" block=SQ event=265 descr="Number of data write requests to the TC (No-Masking, nondeterministic)"></metric>
<metric name="SQC_TC_DATA_ATOMIC_REQ" block=SQ event=266 descr="Number of data atomic requests to the TC (No-Masking, nondeterministic)"></metric>
<metric name="SQC_TC_STALL" block=SQ event=267 descr="Valid request stalled TC request interface (no-credits). (No-Masking, nondeterministic, unwindowed)"></metric>
<metric name="SQC_ICACHE_BUSY_CYCLES" block=SQ event=269 descr="Clock cycles while cache is reporting that it is busy. (No-Masking, nondeterministic, unwindowed)"></metric>
<metric name="SQC_ICACHE_REQ" block=SQ event=270 descr="Number of requests. (per-SQ, per-Bank)"></metric>
<metric name="SQC_ICACHE_HITS" block=SQ event=271 descr="Number of cache hits. (per-SQ, per-Bank, nondeterministic)"></metric>
<metric name="SQC_ICACHE_MISSES" block=SQ event=272 descr="Number of cache misses, includes uncached requests. (per-SQ, per-Bank, nondeterministic)"></metric>
<metric name="SQC_ICACHE_MISSES_DUPLICATE" block=SQ event=273 descr="Number of misses that were duplicates (access to a non-resident, miss pending CL). (per-SQ, per-Bank, nondeterministic)" ></metric>
<metric name="SQC_DCACHE_BUSY_CYCLES" block=SQ event=289 descr=" Clock cycles while cache is reporting that it is busy. (No-Masking, nondeterministic, unwindowed)"></metric>
<metric name="SQC_DCACHE_REQ" block=SQ event=290 descr="Number of requests (post-bank-serialization). (per-SQ, per-Bank)"></metric>
<metric name="SQC_DCACHE_HITS" block=SQ event=291 descr="Number of cache hits. (per-SQ, per-Bank, nondeterministic)"></metric>
<metric name="SQC_DCACHE_MISSES" block=SQ event=292 descr="Number of cache misses, includes uncached requests. (per-SQ, per-Bank, nondeterministic)"></metric>
@@ -591,6 +594,7 @@
<metric name="TCC_RW_REQ" block=TCC event=8 descr="The number of RW requests. This is measured at the tag block."></metric>
<metric name="TCC_PROBE" block=TCC event=9 descr="Number of probe requests. Not windowable."></metric>
<metric name="TCC_PROBE_ALL" block=TCC event=10 descr="Number of external probe requests with with EA_TCC_preq_all== 1. Not windowable."></metric>
<metric name="TCC_INTERNAL_PROBE" block=TCC event=11 descr="Number of self-probes spawned by TCC for CC writes/atomic operations. Not windowable."></metric>
<metric name="TCC_READ" block=TCC event=12 descr="Number of read requests. Compressed reads are included in this, but metadata reads are not included."></metric>
<metric name="TCC_WRITE" block=TCC event=13 descr="Number of write requests."></metric>
<metric name="TCC_ATOMIC" block=TCC event=14 descr="Number of atomic requests of all types."></metric>
@@ -599,6 +603,7 @@
<metric name="TCC_WRITEBACK" block=TCC event=22 descr="Number of lines written back to main memory. This includes writebacks of dirty lines and uncached write/atomic requests."></metric>
<metric name="TCC_EA0_WRREQ" block=TCC event=26 descr="Number of transactions (either 32-byte or 64-byte) going over the TC_EA_wrreq interface. Atomics may travel over the same interface and are generally classified as write requests. This does not include probe commands."></metric>
<metric name="TCC_EA0_WRREQ_64B" block=TCC event=27 descr="Number of 64-byte transactions going (64-byte write or CMPSWAP) over the TC_EA_wrreq interface."></metric>
<metric name="TCC_EA0_WRREQ_PROBE_COMMAND" block=TCC event=28 descr="Number of probe commands going over the TC_EA_wrreq interface."></metric>
<metric name="TCC_EA0_WR_UNCACHED_32B" block=TCC event=29 descr="Number of 32-byte write/atomic going over the TC_EA_wrreq interface due to uncached traffic. Note that CC mtypes can produce uncached requests, and those are included in this. A 64-byte request will be counted as 2"></metric>
<metric name="TCC_EA0_WRREQ_STALL" block=TCC event=30 descr="Number of cycles a write request was stalled."></metric>
<metric name="TCC_EA0_WRREQ_IO_CREDIT_STALL" block=TCC event=31 descr="Number of cycles a EA write request was stalled because the interface was out of IO credits."></metric>
@@ -620,6 +625,7 @@
<metric name="TCC_ALL_TC_OP_WB_WRITEBACK" block=TCC event=73 descr="Number of writebacks due to all TC_OP writeback requests."></metric>
<metric name="TCC_NORMAL_EVICT" block=TCC event=74 descr="Number of evictions due to requests that are not invalidate or probe requests."></metric>
<metric name="TCC_ALL_TC_OP_INV_EVICT" block=TCC event=80 descr="Number of evictions due to all TC_OP invalidate requests."></metric>
<metric name="TCC_PROBE_EVICT" block=TCC event=81 descr="Number of evictions/invalidations due to probes. Not windowable."></metric>
<metric name="TCC_EA0_RDREQ_DRAM" block=TCC event=102 descr="Number of TCC/EA read requests (either 32-byte or 64-byte) destined for DRAM (MC)."></metric>
<metric name="TCC_EA0_WRREQ_DRAM" block=TCC event=103 descr="Number of TCC/EA write requests (either 32-byte of 64-byte) destined for DRAM (MC)."></metric>
</gfx940>
projects/rocprofiler/src/core/counters/metrics/derived_counters.xml
+10 -8
Ver Arquivo
@@ -49,8 +49,8 @@
<metric name="TCC_EA_WRREQ_64B_sum" expr=sum(TCC_EA_WRREQ_64B,16) descr="Number of 64-byte transactions going (64-byte write or CMPSWAP) over the TC_EA_wrreq interface. Sum over TCC instances."></metric>
<metric name="TCC_WRREQ_STALL_max" expr=max(TCC_EA_WRREQ_STALL,16) descr="Number of cycles a write request was stalled. Max over TCC instances."></metric>
<metric name="TCP_TCP_TA_DATA_STALL_CYCLES_sum" expr=sum(TCP_TCP_TA_DATA_STALL_CYCLES,16) decr="Total number of TCP stalls TA data interface."></metric>
<metric name="TCP_TCP_TA_DATA_STALL_CYCLES_max" expr=max(TCP_TCP_TA_DATA_STALL_CYCLES,16) decr="Maximum number of TCP stalls TA data interface."></metric>
<metric name="TCP_TCP_TA_DATA_STALL_CYCLES_sum" expr=sum(TCP_TCP_TA_DATA_STALL_CYCLES,16) descr="Total number of TCP stalls TA data interface."></metric>
<metric name="TCP_TCP_TA_DATA_STALL_CYCLES_max" expr=max(TCP_TCP_TA_DATA_STALL_CYCLES,16) descr="Maximum number of TCP stalls TA data interface."></metric>
<metric name="FETCH_SIZE" expr=(TCC_EA_RDREQ_32B_sum*32+(TCC_EA_RDREQ_sum-TCC_EA_RDREQ_32B_sum)*64)/1024 descr="The total kilobytes fetched from the video memory. This is measured with all extra fetches and any cache or memory effects taken into account."></metric>
<metric name="WRITE_SIZE" expr=((TCC_EA_WRREQ_sum-TCC_EA_WRREQ_64B_sum)*32+TCC_EA_WRREQ_64B_sum*64)/1024 descr="The total kilobytes written to the video memory. This is measured with all extra fetches and any cache or memory effects taken into account."></metric>
@@ -291,8 +291,8 @@
</gfx90a_expr>
<gfx940_expr>
<metric name="TCP_TCP_TA_DATA_STALL_CYCLES_sum" expr=sum(TCP_TCP_TA_DATA_STALL_CYCLES,16) decr="Total number of TCP stalls TA data interface."></metric>
<metric name="TCP_TCP_TA_DATA_STALL_CYCLES_max" expr=max(TCP_TCP_TA_DATA_STALL_CYCLES,16) decr="Maximum number of TCP stalls TA data interface."></metric>
<metric name="TCP_TCP_TA_DATA_STALL_CYCLES_sum" expr=sum(TCP_TCP_TA_DATA_STALL_CYCLES,16) descr="Total number of TCP stalls TA data interface."></metric>
<metric name="TCP_TCP_TA_DATA_STALL_CYCLES_max" expr=max(TCP_TCP_TA_DATA_STALL_CYCLES,16) descr="Maximum number of TCP stalls TA data interface."></metric>
<metric name="VFetchInsts" expr=(SQ_INSTS_VMEM_RD-TA_FLAT_READ_WAVEFRONTS_sum)/SQ_WAVES descr="The average number of vector fetch instructions from the video memory executed per work-item (affected by flow control). Excludes FLAT instructions that fetch from video memory."></metric>
<metric name="VWriteInsts" expr=(SQ_INSTS_VMEM_WR-TA_FLAT_WRITE_WAVEFRONTS_sum)/SQ_WAVES descr="The average number of vector write instructions to the video memory executed per work-item (affected by flow control). Excludes FLAT instructions that write to video memory."></metric>
@@ -369,7 +369,7 @@
<metric name="TCA_CYCLE_sum" expr=sum(TCA_CYCLE,16) descr="Number of cycles. Sum over all TCA instances "></metric>
<metric name="TCA_BUSY_sum" expr=sum(TCA_BUSY,16) descr="Number of cycles we have a request pending. Sum over all TCA instances."></metric>
<metric name="TCC_BUSY_avr" expr=avr(TCC_BUSY,32) descr="TCC_BUSY avr over all memory channels."></metric>
<metric name="TCC_WRREQ_STALL_max" expr=max(TCC_EA_WRREQ_STALL,32) descr="Number of cycles a write request was stalled. Max over TCC instances."></metric>
<metric name="TCC_WRREQ_STALL_max" expr=max(TCC_EA0_WRREQ_STALL,32) descr="Number of cycles a write request was stalled. Max over TCC instances."></metric>
<metric name="TCC_CYCLE_sum" expr=sum(TCC_CYCLE,32) descr="Number of cycles. Not windowable. Sum over TCC instances."></metric>
<metric name="TCC_BUSY_sum" expr=sum(TCC_BUSY,32) descr="Number of cycles we have a request pending. Not windowable. Sum over TCC instances."></metric>
<metric name="TCC_REQ_sum" expr=sum(TCC_REQ,32) descr="Number of requests of all types. This is measured at the tag block. This may be more than the number of requests arriving at the TCC, but it is a good indication of the total amount of work that needs to be performed. Sum over TCC instances."></metric>
@@ -411,9 +411,9 @@
<metric name="TCC_ALL_TC_OP_INV_EVICT_sum" expr=sum(TCC_ALL_TC_OP_INV_EVICT,32) descr="Number of evictions due to all TC_OP invalidate requests. Sum over TCC instances."></metric>
<metric name="TCC_EA0_RDREQ_DRAM_sum" expr=sum(TCC_EA0_RDREQ_DRAM,32) descr="Number of TCC/EA read requests (either 32-byte or 64-byte) destined for DRAM (MC). Sum over TCC instances."></metric>
<metric name="TCC_EA0_WRREQ_DRAM_sum" expr=sum(TCC_EA0_WRREQ_DRAM,32) descr="Number of TCC/EA write requests (either 32-byte of 64-byte) destined for DRAM (MC). Sum over TCC instances."></metric>
<metric name="FETCH_SIZE" expr=(TCC_EA0_RDREQ_32B_sum*32+(TCC_EA_RDREQ_sum-TCC_EA_RDREQ_32B_sum)*64)/1024 descr="The total kilobytes fetched from the video memory. This is measured with all extra fetches and any cache or memory effects taken into account."></metric>
<metric name="WRITE_SIZE" expr=((TCC_EA0_WRREQ_sum-TCC_EA_WRREQ_64B_sum)*32+TCC_EA_WRREQ_64B_sum*64)/1024 descr="The total kilobytes written to the video memory. This is measured with all extra fetches and any cache or memory effects taken into account."></metric>
<metric name="WRITE_REQ_32B" expr=TCC_EA0_WRREQ_64B_sum*2+(TCC_EA_WRREQ_sum-TCC_EA_WRREQ_64B_sum) descr="The total number of 32-byte effective memory writes."></metric>
<metric name="FETCH_SIZE" expr=(TCC_EA0_RDREQ_32B_sum*32+(TCC_EA0_RDREQ_sum-TCC_EA0_RDREQ_32B_sum)*64)/1024 descr="The total kilobytes fetched from the video memory. This is measured with all extra fetches and any cache or memory effects taken into account."></metric>
<metric name="WRITE_SIZE" expr=((TCC_EA0_WRREQ_sum-TCC_EA0_WRREQ_64B_sum)*32+TCC_EA0_WRREQ_64B_sum*64)/1024 descr="The total kilobytes written to the video memory. This is measured with all extra fetches and any cache or memory effects taken into account."></metric>
<metric name="WRITE_REQ_32B" expr=TCC_EA0_WRREQ_64B_sum*2+(TCC_EA0_WRREQ_sum-TCC_EA0_WRREQ_64B_sum) descr="The total number of 32-byte effective memory writes."></metric>
<metric name="CU_OCCUPANCY" expr=(SQ_CYCLES/(SQ_WAVE_CYCLES*4))/MAX_WAVE_SIZE descr="The ratio of active waves on a CU to the maximum number of active waves supported by the CU"></metric>
<metric name="CU_UTILIZATION" expr=GRBM_GUI_ACTIVE/GRBM_COUNT descr="The total number of active cycles divided by total number of elapsed cycles"></metric>
<metric name="TOTAL_16_OPS" expr=(SQ_INSTS_VALU_FMA_F16*2+SQ_INSTS_VALU_ADD_F16+SQ_INSTS_VALU_MUL_F16+SQ_INSTS_VALU_TRANS_F16)*64+((SQ_INSTS_VALU_MFMA_MOPS_F16+SQ_INSTS_VALU_MFMA_MOPS_BF16)*512) descr="The number of 16 bits OPS executed"></metric>
@@ -503,6 +503,8 @@
<gfx908 base="gfx908_expr"></gfx908>
# Aldebaran
<gfx90a base="gfx90a_expr"></gfx90a>
#Mi300
<gfx940 base="gfx940_expr"></gfx940>
#Navi21
<gfx1030 base="gfx1030_expr"></gfx1030>
<gfx1031 base="gfx1031_expr"></gfx1031>
projects/rocprofiler/src/core/counters/metrics/metrics.h
+7 -3
Ver Arquivo
@@ -203,10 +203,14 @@ class MetricsDict {
xml_->AddConst("top.const.metric", "SE_NUM", agent_info->getShaderEngineCount());
ImportMetrics(agent_info, "const");
agent_name_ = agent_info->getName();
if (agent_name_.substr(0, 6) == "gfx940")
agent_name_ =
"gfx940"; // To correct the agent_name from "gfx940:forcestoresc1+" -> "gfx940"
if (std::string("gfx906") == agent_name_ || std::string("gfx908") == agent_name_ ||
std::string("gfx90a") == agent_name_ || std::string("gfx1032") == agent_name_ ||
std::string("gfx1031") == agent_name_ || std::string("gfx1030") == agent_name_ ||
std::string("gfx1100") == agent_name_ || std::string("gfx1101") == agent_name_) {
std::string("gfx90a") == agent_name_ || std::string("gfx940") == agent_name_ ||
std::string("gfx1032") == agent_name_ || std::string("gfx1031") == agent_name_ ||
std::string("gfx1030") == agent_name_ || std::string("gfx1100") == agent_name_ ||
std::string("gfx1101") == agent_name_) {
ImportMetrics(agent_info, agent_name_);
} else {
agent_name_ = agent_info->getGfxip();
projects/rocprofiler/src/core/metrics.h
+42 -26
Ver Arquivo
@@ -136,7 +136,8 @@ class MetricsDict {
const Metric* metric = NULL;
auto it = cache_.find(name);
if (it != cache_.end()) metric = it->second;
if (it != cache_.end())
metric = it->second;
else {
const std::size_t pos = name.find(':');
if (pos != std::string::npos) {
@@ -148,7 +149,8 @@ class MetricsDict {
const std::size_t pos1 = block_name.find('[');
if (pos1 != std::string::npos) {
const std::size_t pos2 = block_name.find(']');
if (pos2 == std::string::npos) EXC_RAISING(HSA_STATUS_ERROR, "Malformed metric name '" << name << "'");
if (pos2 == std::string::npos)
EXC_RAISING(HSA_STATUS_ERROR, "Malformed metric name '" << name << "'");
block_name = name.substr(0, pos1);
const std::string block_index_str = name.substr(pos1 + 1, pos2 - (pos1 + 1));
block_index = atol(block_index_str.c_str());
@@ -156,8 +158,10 @@ class MetricsDict {
}
const hsa_ven_amd_aqlprofile_id_query_t query = Translate(agent_info_, block_name);
const hsa_ven_amd_aqlprofile_block_name_t block_id = (hsa_ven_amd_aqlprofile_block_name_t)query.id;
if ((query.instance_count > 1) && (indexed == false)) EXC_RAISING(HSA_STATUS_ERROR, "Malformed indexed metric name '" << name << "'");
const hsa_ven_amd_aqlprofile_block_name_t block_id =
(hsa_ven_amd_aqlprofile_block_name_t)query.id;
if ((query.instance_count > 1) && (indexed == false))
EXC_RAISING(HSA_STATUS_ERROR, "Malformed indexed metric name '" << name << "'");
const uint32_t event_id = atol(event_str.c_str());
const counter_t counter = {name, {block_id, block_index, event_id}};
metric = new BaseMetric(name, counter);
@@ -189,22 +193,23 @@ class MetricsDict {
const char* xml_name = getenv("ROCP_METRICS");
if (xml_name != NULL) {
xml_ = xml::Xml::Create(xml_name);
if (xml_ == NULL) EXC_RAISING(HSA_STATUS_ERROR, "metrics .xml open error '" << xml_name << "'");
if (xml_ == NULL)
EXC_RAISING(HSA_STATUS_ERROR, "metrics .xml open error '" << xml_name << "'");
xml_->AddConst("top.const.metric", "MAX_WAVE_SIZE", agent_info->max_wave_size);
xml_->AddConst("top.const.metric", "CU_NUM", agent_info->cu_num);
xml_->AddConst("top.const.metric", "SIMD_NUM", agent_info->simds_per_cu * agent_info->cu_num);
xml_->AddConst("top.const.metric", "SE_NUM", agent_info->se_num);
ImportMetrics(agent_info, "const");
agent_name_ = agent_info->name;
if (std::string("gfx906") == agent_info->name ||
std::string("gfx908") == agent_info->name ||
std::string("gfx90a") == agent_info->name ||
std::string("gfx1032") == agent_info->name ||
std::string("gfx1031") == agent_info->name ||
std::string("gfx1030") == agent_info->name ||
std::string("gfx1100") == agent_info->name ||
std::string("gfx1101") == agent_info->name) {
ImportMetrics(agent_info, agent_info->name);
if (agent_name_.substr(0, 6) == "gfx940")
agent_name_ =
"gfx940"; // To correct the agent_name from "gfx940:forcestoresc1+" -> "gfx940"
if (std::string("gfx906") == agent_name_ || std::string("gfx908") == agent_name_ ||
std::string("gfx90a") == agent_name_ || std::string("gfx940") == agent_name_ ||
std::string("gfx1032") == agent_name_ || std::string("gfx1031") == agent_name_ ||
std::string("gfx1030") == agent_name_ || std::string("gfx1100") == agent_name_ ||
std::string("gfx1101") == agent_name_) {
ImportMetrics(agent_info, agent_name_);
} else {
agent_name_ = agent_info->gfxip;
ImportMetrics(agent_info, agent_info->gfxip);
@@ -218,14 +223,16 @@ class MetricsDict {
for (auto& entry : cache_) delete entry.second;
}
static hsa_ven_amd_aqlprofile_id_query_t Translate(const util::AgentInfo* agent_info, const std::string& block_name) {
static hsa_ven_amd_aqlprofile_id_query_t Translate(const util::AgentInfo* agent_info,
const std::string& block_name) {
hsa_ven_amd_aqlprofile_profile_t profile{};
profile.agent = agent_info->dev_id;
hsa_ven_amd_aqlprofile_id_query_t query = {block_name.c_str(), 0, 0};
hsa_status_t status =
util::HsaRsrcFactory::Instance().AqlProfileApi()->hsa_ven_amd_aqlprofile_get_info(
&profile, HSA_VEN_AMD_AQLPROFILE_INFO_BLOCK_ID, &query);
if (status != HSA_STATUS_SUCCESS) AQL_EXC_RAISING(HSA_STATUS_ERROR, "ImportMetrics: bad block name '" << block_name << "'");
if (status != HSA_STATUS_SUCCESS)
AQL_EXC_RAISING(HSA_STATUS_ERROR, "ImportMetrics: bad block name '" << block_name << "'");
return query;
}
@@ -251,7 +258,8 @@ class MetricsDict {
const uint32_t event_id = atol(event_str.c_str());
const hsa_ven_amd_aqlprofile_id_query_t query = Translate(agent_info, block_name);
const hsa_ven_amd_aqlprofile_block_name_t block_id = (hsa_ven_amd_aqlprofile_block_name_t)query.id;
const hsa_ven_amd_aqlprofile_block_name_t block_id =
(hsa_ven_amd_aqlprofile_block_name_t)query.id;
if (query.instance_count > 1) {
for (unsigned block_index = 0; block_index < query.instance_count; ++block_index) {
std::ostringstream full_name;
@@ -272,9 +280,11 @@ class MetricsDict {
xml::Expr* expr_obj = NULL;
try {
expr_obj = new xml::Expr(expr_str, new ExprCache(&cache_));
} catch(const xml::exception_t& exc) {
if (do_lookup) metrics_list.push_back(node);
else throw(exc);
} catch (const xml::exception_t& exc) {
if (do_lookup)
metrics_list.push_back(node);
else
throw(exc);
}
if (expr_obj) {
#if 0
@@ -285,7 +295,8 @@ class MetricsDict {
for (const std::string& var : expr_obj->GetVars()) {
auto it = cache_.find(var);
if (it == cache_.end()) {
EXC_RAISING(HSA_STATUS_ERROR, "Bad metric '" << name << "', var '" << var << "' is not found");
EXC_RAISING(HSA_STATUS_ERROR,
"Bad metric '" << name << "', var '" << var << "' is not found");
}
it->second->GetCounters(counters_vec);
}
@@ -309,23 +320,27 @@ class MetricsDict {
}
}
const Metric* AddMetric(const std::string& name, const std::string& /*alias*/, const counter_t& counter) {
const Metric* AddMetric(const std::string& name, const std::string& /*alias*/,
const counter_t& counter) {
const Metric* metric = NULL;
const auto ret = cache_.insert({name, NULL});
if (ret.second) {
metric = new BaseMetric(name, counter);
ret.first->second = metric;
} else EXC_RAISING(HSA_STATUS_ERROR, "metric redefined '" << name << "'");
} else
EXC_RAISING(HSA_STATUS_ERROR, "metric redefined '" << name << "'");
return metric;
}
const Metric* AddMetric(const std::string& name, const counters_vec_t& counters_vec, const xml::Expr* expr_obj) {
const Metric* AddMetric(const std::string& name, const counters_vec_t& counters_vec,
const xml::Expr* expr_obj) {
const Metric* metric = NULL;
const auto ret = cache_.insert({name, NULL});
if (ret.second) {
metric = new ExprMetric(name, counters_vec, expr_obj);
ret.first->second = metric;
} else EXC_RAISING(HSA_STATUS_ERROR, "expr-metric redefined '" << name << "'");
} else
EXC_RAISING(HSA_STATUS_ERROR, "expr-metric redefined '" << name << "'");
return metric;
}
@@ -336,7 +351,8 @@ class MetricsDict {
printf("> Metric '%s'\n", metric->GetName().c_str());
metric->GetCounters(counters_vec);
for (auto c : counters_vec) {
printf(" counter %s, b(%u), i (%u), e (%u)\n", c->name.c_str(), c->event.block_name, c->event.block_index, c->event.counter_id);
printf(" counter %s, b(%u), i (%u), e (%u)\n", c->name.c_str(), c->event.block_name,
c->event.block_index, c->event.counter_id);
}
}
}